mctp: perform route lookups under a RCU read-side lock

Our current route lookups (mctp_route_lookup and mctp_route_lookup_null)
traverse the net's route list without the RCU read lock held. This means
the route lookup is subject to preemption, resulting in an potential
grace period expiry, and so an eventual kfree() while we still have the
route pointer.

Add the proper read-side critical section locks around the route
lookups, preventing premption and a possible parallel kfree.

The remaining net->mctp.routes accesses are already under a
rcu_read_lock, or protected by the RTNL for updates.

Based on an analysis from Sili Luo <[e-mail]>, where
introducing a delay in the route lookup could cause a UAF on
simultaneous sendmsg() and route deletion.

Reported-by: Sili Luo <[e-mail]>
Fixes: 889b7da23abf ("mctp: Add initial routing framework")
Cc: [e-mail]
Signed-off-by: Jeremy Kerr <[e-mail]>
Reviewed-by: Eric Dumazet <[e-mail]>
Link: https://lore.kernel.org/r/[e-mail]
Signed-off-by: Jakub Kicinski <[e-mail]>

net/mctp/route.c

···
 {
 	struct mctp_route *tmp, *rt = NULL;
 
+	rcu_read_lock();
+
 	list_for_each_entry_rcu(tmp, &net->mctp.routes, list) {
 		/* TODO: add metrics */
 		if (mctp_rt_match_eid(tmp, dnet, daddr)) {
···
 		}
 	}
 
+	rcu_read_unlock();
+
 	return rt;
 }
 
 static struct mctp_route *mctp_route_lookup_null(struct net *net,
 						 struct net_device *dev)
 {
-	struct mctp_route *rt;
+	struct mctp_route *tmp, *rt = NULL;
 
-	list_for_each_entry_rcu(rt, &net->mctp.routes, list) {
-		if (rt->dev->dev == dev && rt->type == RTN_LOCAL &&
-		    refcount_inc_not_zero(&rt->refs))
-			return rt;
+	rcu_read_lock();
+
+	list_for_each_entry_rcu(tmp, &net->mctp.routes, list) {
+		if (tmp->dev->dev == dev && tmp->type == RTN_LOCAL &&
+		    refcount_inc_not_zero(&tmp->refs)) {
+			rt = tmp;
+			break;
+		}
 	}
 
-	return NULL;
+	rcu_read_unlock();
+
+	return rt;
 }
 
 static int mctp_do_fragment_route(struct mctp_route *rt, struct sk_buff *skb,